Macrophages are phagocytic cells that recognize and 'eat' foreign cellular microbes, as well as dying cells and aberrant cells such as cancers. They display a variety of receptors by which they recognize 'eat me' and 'don't eat me' signals on their target cells. Macrophages inspect migrating hematopoietic stem cells (HSC) as they pass through the sinusoids. Macrophages also inspect cancer cells, which express pro-phagocytic 'eat me' signals to ensure that aberrant cells are ingested in a process termed programmed cell removal. However, successful cancer cells evade programmed cell removal by expressing CD47, a dominant 'don't eat me' signal that is a ligand for SIRP, an inhibitory receptor expressed on macrophages. Ligation of SIRP by CD47 blocks macrophages from phagocytosing the tumor cells. The CD47/SIRP axis represents a toggle switch that can be blocked or stimulated for different therapeutic goals. Blocking binding of CD47 to SIRP promotes phagocytosis of cancer cells. Conversely, the CD47/SIRP interaction is a critical determinant of engraftment success in hematopoietic cell transplantation (HCT): CD47 expression levels on HSC correlate with their relative engraftability. Unfortunately, bone marrow HSC express low levels of CD47, accounting for their relative inefficiency in transplantation. Thus, depending on the clinical scenario, the SIRP/CD47 system offers an exciting new axis for both anti-tumor and transplant therapy. However, effective utilization of the soluble ectodomains of either SIRP or CD47 as agonists, or antagonists is limited by the low affinity of the wild-type CD47/SIRP interaction. We wish to execute a structure-based engineering approach to creating high-affinity SIRP and CD47 ectodomains as therapeutics for cancer and HCT. We propose to combine the expertise of the Garcia lab in structural biology, protein engineering and immune intervention, with the strengths of the Weissman and Shizuru labs in in vivo cancer, and hematopoietic cell transplantation biology, to target the CD47/SIRP axis. For Aim #1, we have engineered high-affinity SIRP monomers, that prevent the interaction between endogenous CD47 and SIRP, and dramatically synergize with clinically-established therapeutic monoclonal antibodies in stimulating phagocytosis of tumor cells in vivo. This 1-2 punch of target sensitization to macrophages followed by cytolysis by anti-tumor mAb is a completely novel strategy. For Aim #2, we are engineering high-affinity CD47 variants that activate SIRP inhibitory signaling and decrease macrophage activation, thus enhancing HSC engraftment. Finally, in Aim #3 we wish to reconstitute and molecularly characterize the cell surface 'don't eat me' complex composed of CD47 and SIRP with putative alternative ligands including thrombospondin and integrins in order to fully understand the therapeutic potential of this system for modulating macrophage phagocytosis. We anticipate these studies will yield novel classes of biotherapeutics with applications in many types of human cancers, and for transplant therapy.